Open taylorreiter opened 2 years ago
what was the rule and/or command line?
thanks :)
rule bcalm_catlas_prepare_input:
Invoked with
python -m spacegraphcats build {input.conf} --outdir={params.outdir} --rerun-incomplete --nolock
With conf file
catlas_base: GCF_000765235.1
input_sequences:
- outputs/sgc_genome_queries_hardtrim/GCF_000765235.1.hardtrim.fa.gz
radius: 10
paired_reads: true
ok. I think the problem is that the database should be deleted before that rule is run.
I was able to replicate with:
python -m spacegraphcats build dory-test --outdir=dory-foo --rerun-incomplete --nolock
touch dory-foo/dory_k21/bcalm.inputlist.txt
python -m spacegraphcats build dory-test --outdir=dory-foo --rerun-incomplete --nolock
I'm 99% certain that what's going on is that the output db = f"{cdbg_dir}/bcalm.unitigs.db"
already exists and is explicitly marked as protected
, so snakemake isn't overwriting it.
The short-term fix is to remove that file.
The long-term fix might be to remove protected status, but ...I'm resistant, because I think I put this in place to avoid Very Large Files being overwritten too easily by sgc. So it might be better to have this be a hiccough that you need to manually overcome.
Yes plz don't remove write protections here. I think we should leave this issue open so a solution is searchable. Indeed, I deleted the file and restarted spacegraphcats and it ran perfectly.